Derivatives of a Function in Parametric Form Explained

#Calculus
TL;DR
Derivatives of a function in parametric form come from one ratio: dydx=dy/dtdx/dt\frac{dy}{dx} = \frac{dy/dt}{dx/dt}. This guide covers why it works, the second derivative trap, horizontal and vertical tangents, and six verified examples.
BT
Bhanzu TeamLast updated on October 9, 202617 min read

What Is The Derivative Of A Function In Parametric Form?

The derivative of a function in parametric form is the slope dydx\frac{dy}{dx} of a curve whose points are given by two equations, x=x(t)x = x(t) and y=y(t)y = y(t), in terms of a third variable tt called the parameter. You find it by differentiating both equations with respect to tt and dividing:

dydx=dydtdxdt,dxdt≠0\frac{dy}{dx} = \frac{\dfrac{dy}{dt}}{\dfrac{dx}{dt}}, \qquad \frac{dx}{dt} \neq 0

Here dydt\frac{dy}{dt} is how fast yy changes as tt grows, dxdt\frac{dx}{dt} is how fast xx changes, and their ratio is the change in yy per unit change in xx. The formula needs two conditions: both x(t)x(t) and y(t)y(t) must be differentiable at the value of tt you care about, and dxdt\frac{dx}{dt} must not be zero there. When dxdt=0\frac{dx}{dt} = 0 the formula breaks, and that is exactly where vertical tangents and cusps live.

The equations themselves, how to sketch them and how to convert them, have their own page on parametric equations. This page is about the calculus that sits on top: the first derivative, the second derivative, and the tangents they produce. All derivatives here are written in Leibniz notation, so the variable each one is taken with respect to is always visible.

Why Does The Parametric Formula Work?

The parametric formula is the chain rule solved for the slope. Suppose that near some value of tt, the curve can be read as yy depending on xx, and xx depending on tt. The chain rule then says:

dydt=dydx⋅dxdt\frac{dy}{dt} = \frac{dy}{dx} \cdot \frac{dx}{dt}

Divide both sides by dxdt\frac{dx}{dt}, which is allowed only when it is not zero:

dydx=dy/dtdx/dt\frac{dy}{dx} = \frac{dy/dt}{dx/dt}

The condition dxdt≠0\frac{dx}{dt} \neq 0 is doing real work here, not just avoiding division by zero. If dxdt\frac{dx}{dt} is not zero, then xx is strictly increasing or strictly decreasing in tt near that point, so each xx comes from exactly one tt and yy really is a function of xx there. That local reversal of tt and xx is the same idea behind inverse functions and differentiation, where dtdx=1dx/dt\frac{dt}{dx} = \frac{1}{dx/dt}.

So the full recipe is three lines:

  1. Differentiate x(t)x(t) to get dxdt\frac{dx}{dt}.

  2. Differentiate y(t)y(t) to get dydt\frac{dy}{dt}.

  3. Divide the second by the first, and simplify.

What Does The Parametric Derivative Mean On A Curve?

On a parametric curve, dydx\frac{dy}{dx} is the slope of the tangent line, while the pair (dxdt,dydt)\left(\frac{dx}{dt}, \frac{dy}{dt}\right) is the velocity of the moving point. A question on Quora asks exactly this: what is the interpretation of dydx\frac{dy}{dx} in parametric equations, and why is it different from the velocity?

The answer is that the velocity is an arrow and the slope is only its direction. Picture tt as time and the point (x(t),y(t))(x(t), y(t)) as a car on a road:

  • The velocity vector (dxdt,dydt)\left(\frac{dx}{dt}, \frac{dy}{dt}\right) points along the road and its length is the car's speed.

  • The slope dydx\frac{dy}{dx} is the rise over run of that arrow. It tells you which way the road tilts, but not how fast the car is going.

  • Same road, different speed: trace the curve twice as fast (replace tt by 2t2t) and both components of velocity double, but their ratio, the slope, stays the same.

That last point is a useful check on any answer. The slope belongs to the curve, so it cannot depend on how quickly you travel along it. The full motion story, with speed and acceleration, is on the velocity and acceleration page.

Why Not Just Eliminate The Parameter?

Eliminating the parameter means solving one equation for tt and substituting into the other, so the curve becomes an ordinary y=f(x)y = f(x). Brainly carries the textbook version of the instruction: "if x and y are connected parametrically by the given equations, then without eliminating the parameter find dydx\frac{dy}{dx}". There are three good reasons the parametric formula is usually the better route.

  • Elimination is often impossible in closed form. For the cycloid, x=r(θ−sin⁡θ)x = r(\theta - \sin\theta) cannot be solved for θ\theta with ordinary functions, so there is no tidy y=f(x)y = f(x) to differentiate.

  • Elimination can split one curve into pieces. The unit circle x=cos⁡tx = \cos t, y=sin⁡ty = \sin t becomes y=±1−x2y = \pm\sqrt{1 - x^2}, two functions with a sign choice to track. The parametric slope −cos⁡tsin⁡t-\frac{\cos t}{\sin t} handles the whole circle at once.

  • The answer is wanted in terms of tt anyway. Tangent problems usually name a point by its parameter value, so a slope written in tt is the slope you need.

When elimination is easy, it makes a good check.

Take x=ln⁡tx = \ln t and y=t2y = t^2 with t>0t > 0. Parametrically, dydx=2t1/t=2t2\frac{dy}{dx} = \frac{2t}{1/t} = 2t^2. Eliminating, t=ext = e^x, so y=e2xy = e^{2x} and dydx=2e2x=2t2\frac{dy}{dx} = 2e^{2x} = 2t^2. Both routes agree, as they must.

How Do You Find The Second Derivative In Parametric Form?

The second derivative in parametric form is the derivative of dydx\frac{dy}{dx} with respect to xx, and since dydx\frac{dy}{dx} is written in tt, you differentiate it with respect to tt and divide by dxdt\frac{dx}{dt} once more:

d2ydx2=ddt(dydx)dxdt\frac{d^2y}{dx^2} = \frac{\dfrac{d}{dt}\left(\dfrac{dy}{dx}\right)}{\dfrac{dx}{dt}}

This is the same chain-rule move as before, applied to a new function. Call the first derivative m(t)=dydxm(t) = \frac{dy}{dx}. To find dmdx\frac{dm}{dx}, use dmdt=dmdx⋅dxdt\frac{dm}{dt} = \frac{dm}{dx}\cdot\frac{dx}{dt} and divide. Expanding ddt\frac{d}{dt} of the quotient gives an equivalent one-line form:

d2ydx2=dxdt,d2ydt2−dydt,d2xdt2(dxdt)3\frac{d^2y}{dx^2} = \frac{\dfrac{dx}{dt},\dfrac{d^2y}{dt^2} - \dfrac{dy}{dt},\dfrac{d^2x}{dt^2}}{\left(\dfrac{dx}{dt}\right)^3}

The general ideas behind the second derivative carry over unchanged. A positive value means the curve is concave up at that point (bending upward as xx increases), and a negative value means concave down.

What it is not is d2ydt2\frac{d^2y}{dt^2} divided by d2xdt2\frac{d^2x}{dt^2}. That ratio looks like the natural extension of the first-derivative rule, and it is wrong almost everywhere; Example 3 below lets it fail on a circle, where the right sign is obvious from the picture. Higher orders follow the same pattern: each new derivative is ddt\frac{d}{dt} of the previous one, divided by dxdt\frac{dx}{dt}.

How Do You Find Horizontal And Vertical Tangents?

Horizontal and vertical tangents come straight from the numerator and denominator of the slope formula. Because dydx\frac{dy}{dx} is a fraction, you read its zeros and its blow-ups from the top and the bottom separately:

Condition at t0t_0

What the curve does

Example

dydt=0\frac{dy}{dt} = 0 and dxdt≠0\frac{dx}{dt} \neq 0

Horizontal tangent (slope 0)

Top of a circle

dxdt=0\frac{dx}{dt} = 0 and dydt≠0\frac{dy}{dt} \neq 0

Vertical tangent (slope undefined)

Rightmost point of a circle

dxdt=0\frac{dx}{dt} = 0 and dydt=0\frac{dy}{dt} = 0

No conclusion yet: could be a cusp, a corner or a smooth tangent

Cusp of a cycloid

The third row deserves a careful look. When both rates are zero, the moving point has stopped, and the shape of the curve there depends on how it starts moving again. You settle it by finding the limit of dydx\frac{dy}{dx} as t→t0t \to t_0; for the cycloid that limit is infinite, which gives the sharp, vertical-tipped cusp in the opening picture.

Once you have a slope mm at a parameter value t0t_0, the tangent line through the point (x0,y0)=(x(t0),y(t0))(x_0, y_0) = (x(t_0), y(t_0)) is the usual point-slope line, covered in full on the tangent line equations page:

y−y0=m,(x−x0)y - y_0 = m,(x - x_0)

Examples Of Derivatives Of A Function In Parametric Form

These six examples run from a one-line slope to a curve with two tangents at the same point.

Example 1

Find dydx\frac{dy}{dx} for x=t3+1x = t^3 + 1, y=t2+4ty = t^2 + 4t, and evaluate it at t=1t = 1.

dxdt=3t2\frac{dx}{dt} = 3t^2

dydt=2t+4\frac{dy}{dt} = 2t + 4

dydx=2t+43t2=2(t+2)3t2\frac{dy}{dx} = \frac{2t + 4}{3t^2} = \frac{2(t + 2)}{3t^2}

dydx∣t=1=2(3)3=2\left.\frac{dy}{dx}\right|_{t = 1} = \frac{2(3)}{3} = 2

Final answer: dydx=2(t+2)3t2\frac{dy}{dx} = \frac{2(t + 2)}{3t^2}, which equals 22 at t=1t = 1 (the point (2,5)(2, 5)). At t=0t = 0 the formula is undefined, because dxdt=0\frac{dx}{dt} = 0 there.

Example 2

Find the tangent line to the ellipse x=3cos⁡tx = 3\cos t, y=2sin⁡ty = 2\sin t at t=π4t = \frac{\pi}{4}.

The ellipse here has semi-axes 3 and 2. Differentiate each coordinate:

dxdt=−3sin⁡t,dydt=2cos⁡t\frac{dx}{dt} = -3\sin t, \qquad \frac{dy}{dt} = 2\cos t

dydx=2cos⁡t−3sin⁡t=−23cot⁡t\frac{dy}{dx} = \frac{2\cos t}{-3\sin t} = -\frac{2}{3}\cot t

At t=π4t = \frac{\pi}{4}, cot⁡π4=1\cot\frac{\pi}{4} = 1, so the slope is −23-\frac{2}{3}. The point is:

x0=3cos⁡π4=322,y0=2sin⁡π4=2x_0 = 3\cos\frac{\pi}{4} = \frac{3\sqrt{2}}{2}, \qquad y_0 = 2\sin\frac{\pi}{4} = \sqrt{2}

y−2=−23(x−322)y - \sqrt{2} = -\frac{2}{3}\left(x - \frac{3\sqrt{2}}{2}\right)

y=−23x+22y = -\frac{2}{3}x + 2\sqrt{2}

Check with the standard ellipse tangent xx09+yy04=1\frac{x x_0}{9} + \frac{y y_0}{4} = 1: it gives 2x+3y=622x + 3y = 6\sqrt{2}, the same line.

Final answer: y=−23x+22y = -\frac{2}{3}x + 2\sqrt{2}.

Example 3

For the unit circle x=cos⁡(πt)x = \cos(\pi t), y=sin⁡(πt)y = \sin(\pi t), find dydx\frac{dy}{dx} and d2ydx2\frac{d^2y}{dx^2}, then evaluate the second derivative at t=14t = \frac{1}{4}.

This is a Brainly question almost word for word, and it is the cleanest place to watch the tempting shortcut fail.

The first derivative is routine:

dxdt=−πsin⁡(πt),dydt=πcos⁡(πt)\frac{dx}{dt} = -\pi\sin(\pi t), \qquad \frac{dy}{dt} = \pi\cos(\pi t)

dydx=πcos⁡(πt)−πsin⁡(πt)=−cot⁡(πt)\frac{dy}{dx} = \frac{\pi\cos(\pi t)}{-\pi\sin(\pi t)} = -\cot(\pi t)

The tempting move for the second derivative is to divide the second derivatives, just as the first derivatives were divided:

d2y/dt2d2x/dt2=−π2sin⁡(πt)−π2cos⁡(πt)=tan⁡(πt)\frac{d^2y/dt^2}{d^2x/dt^2} = \frac{-\pi^2\sin(\pi t)}{-\pi^2\cos(\pi t)} = \tan(\pi t)

At t=14t = \frac{1}{4} this gives tan⁡π4=1\tan\frac{\pi}{4} = 1, a positive number. But t=14t = \frac{1}{4} is the point (22,22)\left(\frac{\sqrt{2}}{2}, \frac{\sqrt{2}}{2}\right) on the upper half of the circle, and the upper half bends downward like a dome. A concave-down curve must have a negative second derivative, so the shortcut has broken.

The correct method differentiates the first derivative with respect to tt, then divides by dxdt\frac{dx}{dt}:

ddt(−cot⁡(πt))=πcsc⁡2(πt)\frac{d}{dt}\left(-\cot(\pi t)\right) = \pi\csc^2(\pi t)

d2ydx2=πcsc⁡2(πt)−πsin⁡(πt)=−csc⁡3(πt)\frac{d^2y}{dx^2} = \frac{\pi\csc^2(\pi t)}{-\pi\sin(\pi t)} = -\csc^3(\pi t)

d2ydx2∣t=1/4=−(2)3=−22\left.\frac{d^2y}{dx^2}\right|_{t = 1/4} = -\left(\sqrt{2}\right)^3 = -2\sqrt{2}

Check with implicit differentiation on x2+y2=1x^2 + y^2 = 1, which gives d2ydx2=−1y3\frac{d^2y}{dx^2} = -\frac{1}{y^3}. At y=22y = \frac{\sqrt{2}}{2} that is −22-2\sqrt{2} as well.

Final answer: dydx=−cot⁡(πt)\frac{dy}{dx} = -\cot(\pi t) and d2ydx2=−csc⁡3(πt)\frac{d^2y}{dx^2} = -\csc^3(\pi t), which equals −22-2\sqrt{2} at t=14t = \frac{1}{4}.

Example 4

For the curve x=2t3−3t2x = 2t^3 - 3t^2, y=t3−12ty = t^3 - 12t (a Brainly question), find every point with a horizontal or vertical tangent.

dxdt=6t2−6t=6t(t−1)\frac{dx}{dt} = 6t^2 - 6t = 6t(t - 1)

dydt=3t2−12=3(t−2)(t+2)\frac{dy}{dt} = 3t^2 - 12 = 3(t - 2)(t + 2)

Horizontal tangents need dydt=0\frac{dy}{dt} = 0, so t=2t = 2 or t=−2t = -2; at both, dxdt=12\frac{dx}{dt} = 12 or 3636, not zero. Vertical tangents need dxdt=0\frac{dx}{dt} = 0, so t=0t = 0 or t=1t = 1; at both, dydt=−12\frac{dy}{dt} = -12 or −9-9, not zero.

tt

Point (x,y)(x, y)

Tangent

−2-2

(−28,16)(-28, 16)

Horizontal

22

(4,−16)(4, -16)

Horizontal

00

(0,0)(0, 0)

Vertical

11

(−1,−11)(-1, -11)

Vertical

Final answer: horizontal tangents at (−28,16)(-28, 16) and (4,−16)(4, -16); vertical tangents at (0,0)(0, 0) and (−1,−11)(-1, -11).

Example 5

For the cycloid x=r(θ−sin⁡θ)x = r(\theta - \sin\theta), y=r(1−cos⁡θ)y = r(1 - \cos\theta), find dydx\frac{dy}{dx} and d2ydx2\frac{d^2y}{dx^2}, then evaluate both at θ=π2\theta = \frac{\pi}{2} when r=2r = 2.

dxdθ=r(1−cos⁡θ),dydθ=rsin⁡θ\frac{dx}{d\theta} = r(1 - \cos\theta), \qquad \frac{dy}{d\theta} = r\sin\theta

dydx=sin⁡θ1−cos⁡θ=cot⁡θ2\frac{dy}{dx} = \frac{\sin\theta}{1 - \cos\theta} = \cot\frac{\theta}{2}

The last step uses the half-angle identities sin⁡θ=2sin⁡θ2cos⁡θ2\sin\theta = 2\sin\frac{\theta}{2}\cos\frac{\theta}{2} and 1−cos⁡θ=2sin⁡2θ21 - \cos\theta = 2\sin^2\frac{\theta}{2}. For the second derivative:

ddθ(cot⁡θ2)=−12csc⁡2θ2\frac{d}{d\theta}\left(\cot\frac{\theta}{2}\right) = -\frac{1}{2}\csc^2\frac{\theta}{2}

d2ydx2=−12csc⁡2θ2r(1−cos⁡θ)=−1r(1−cos⁡θ)2\frac{d^2y}{dx^2} = \frac{-\frac{1}{2}\csc^2\frac{\theta}{2}}{r(1 - \cos\theta)} = -\frac{1}{r(1 - \cos\theta)^2}

At θ=π2\theta = \frac{\pi}{2} with r=2r = 2, the point is (π−2,2)(\pi - 2, 2), and:

dydx=cot⁡π4=1,d2ydx2=−12(1−0)2=−12\frac{dy}{dx} = \cot\frac{\pi}{4} = 1, \qquad \frac{d^2y}{dx^2} = -\frac{1}{2(1 - 0)^2} = -\frac{1}{2}

The second derivative is negative for every θ\theta between cusps, so each arch is concave down all the way across. At θ=0,2π,4π,…\theta = 0, 2\pi, 4\pi, \ldots both rates are zero and cot⁡θ2\cot\frac{\theta}{2} grows without bound, which is the vertical cusp.

Final answer: dydx=cot⁡θ2\frac{dy}{dx} = \cot\frac{\theta}{2} and d2ydx2=−1r(1−cos⁡θ)2\frac{d^2y}{dx^2} = -\frac{1}{r(1 - \cos\theta)^2}; at the given point the slope is 11 and the second derivative is −12-\frac{1}{2}.

Example 6

The curve x=t2x = t^2, y=t3−3ty = t^3 - 3t passes through (3,0)(3, 0) twice. Find both tangent lines there.

Set y=0y = 0: t(t2−3)=0t(t^2 - 3) = 0, so t=0t = 0 or t=±3t = \pm\sqrt{3}. Only t=±3t = \pm\sqrt{3} give x=3x = 3, so the curve crosses itself at (3,0)(3, 0).

dydx=3t2−32t\frac{dy}{dx} = \frac{3t^2 - 3}{2t}

t=3:9−323=3t = \sqrt{3}: \quad \frac{9 - 3}{2\sqrt{3}} = \sqrt{3}

t=−3:9−3−23=−3t = -\sqrt{3}: \quad \frac{9 - 3}{-2\sqrt{3}} = -\sqrt{3}

Final answer: two tangents at the same point, y=3(x−3)y = \sqrt{3}(x - 3) and y=−3(x−3)y = -\sqrt{3}(x - 3). A graph y=f(x)y = f(x) can never do this, which is one more reason parametric form exists.

Why Do Curves Need A Parameter In The First Place?

Curves need a parameter because many important paths are motions, and a motion is two coordinates changing with time, not one coordinate depending on the other. The cycloid shows why this mattered historically:

  • The curve of quickest descent. In June 1696, Johann Bernoulli challenged mathematicians in Acta Eruditorum to find the path along which a bead slides fastest between two points under gravity. The answer, the brachistochrone curve, is an arc of a cycloid.

  • The clock that keeps time on any swing. Christiaan Huygens proved in Horologium Oscillatorium (1673) that a cycloid is tautochronous: a bead released anywhere on it reaches the bottom in the same time.

  • Slopes without a formula for yy. Both results needed the slope and bending of a curve whose yy cannot be written as a neat function of xx. Parametric differentiation is what makes that question answerable.

The same problem returns every time something moves along a path: a ball in flight, a robot arm, a pen on a plotter. The coordinates are easy to write in time; the slope comes from dividing their rates.

Where Are Parametric Derivatives Used In Real Life?

Parametric derivatives turn up wherever a path is designed or tracked, because the slope of the path is the direction of travel.

  • Projectiles. A ball launched at speed vv and angle α\alpha follows x=vcos⁡α,tx = v\cos\alpha, t, y=vsin⁡α,t−gt22y = v\sin\alpha, t - \frac{g t^2}{2}. Its slope is tan⁡α−gtvcos⁡α\tan\alpha - \frac{g t}{v\cos\alpha}, which reaches zero, the top of the flight, at t=vsin⁡αgt = \frac{v\sin\alpha}{g}.

  • Computer fonts and design. The letters on this screen are drawn with Bézier curves, which are parametric. The tangent direction at each end is what lets two pieces join without a visible kink.

  • Gears and clockwork. Cycloidal tooth profiles, still used in clocks, are shaped so that the tangent directions of meshing teeth match as they roll.

  • Arc length. The length of a parametric path is built from the same two rates, dxdt\frac{dx}{dt} and dydt\frac{dy}{dt}; see arc length of a curve.

Where Solutions Go Off The Rails

Parametric differentiation has very few moving parts, which makes the errors predictable.

Mistake 1: Dividing The Wrong Way Up

Where it slips in:

The very first line of the problem, usually under time pressure.

Don't do this:

Writing dydx=dx/dtdy/dt\frac{dy}{dx} = \frac{dx/dt}{dy/dt}, which gives the reciprocal of the slope.

The correct way:

Keep yy on top on both sides: dydx=dy/dtdx/dt\frac{dy}{dx} = \frac{dy/dt}{dx/dt}. The rusher, who writes the two rates down and divides in whatever order they appeared, is the student this catches; saying "the dtdts cancel" out loud before dividing fixes the order every time.

Mistake 2: Dividing The Second Derivatives

Where it slips in:

Any question that asks for d2ydx2\frac{d^2y}{dx^2} or for concavity.

Don't do this:

Computing d2y/dt2d2x/dt2\frac{d^2y/dt^2}{d^2x/dt^2}. On the circle in Example 3 it gives +1+1 where the true value is −22-2\sqrt{2}.

The correct way:

Differentiate dydx\frac{dy}{dx} with respect to tt, then divide by dxdt\frac{dx}{dt}. The memorizer who learned "divide the tt derivatives" as a rule, rather than as the chain rule, extends it one step too far; the first instinct is to treat the second derivative as the same pattern, and the fix is to remember why the first one worked.

Mistake 3: Stopping One Step Early

Where it slips in:

The second derivative, right after finding ddt(dydx)\frac{d}{dt}\left(\frac{dy}{dx}\right).

Don't do this:

Reporting ddt(dydx)\frac{d}{dt}\left(\frac{dy}{dx}\right) as the answer. For the cycloid that would be −12csc⁡2θ2-\frac{1}{2}\csc^2\frac{\theta}{2}, missing the division by r(1−cos⁡θ)r(1 - \cos\theta).

The correct way:

Divide by dxdt\frac{dx}{dt} again. A quick check: if rr (or any scale) has vanished from a second derivative that should depend on it, a division is missing.

Mistake 4: Calling Every Zero Of dxdt\frac{dx}{dt} A Vertical Tangent

Where it slips in:

Tangent-finding questions on curves with cusps or stopping points.

Don't do this:

Listing every tt with dxdt=0\frac{dx}{dt} = 0 as a vertical tangent without checking dydt\frac{dy}{dt}.

The correct way:

Check both rates. If both are zero, find the limit of dydx\frac{dy}{dx} as tt approaches that value before naming the tangent. The silent understander usually gets the cycloid cusp right by picture but cannot say why; the limit is the reason.

The habit that prevents most of these is writing the two rates on separate lines before combining them. Students who skip straight to the quotient lose track of which rate is which, and the inverted fraction in Mistake 1 is the usual result.

A wrong curve chosen with confidence has a famous real-world version. In 1638, Galileo argued that a circular arc beats any straight chord for quickest descent, which is true, but the fastest path of all turned out to be the cycloid, found only after Bernoulli's 1696 challenge. Trusting the familiar shape over the calculation is the same error as trusting the familiar-looking division in Mistake 2.

Practice Problems On Derivatives Of A Function In Parametric Form

Work each one, then compare with the answers below. Every answer has been recomputed.

  1. Find dydx\frac{dy}{dx} for x=t3x = t^3, y=t2−4ty = t^2 - 4t, and evaluate it at t=1t = 1.

  2. Find dydx\frac{dy}{dx} and d2ydx2\frac{d^2y}{dx^2} for x=etx = e^t, y=tety = t e^t.

  3. Find the points with horizontal and vertical tangents on x=4cos⁡tx = 4\cos t, y=3sin⁡ty = 3\sin t.

  4. Find d2ydx2\frac{d^2y}{dx^2} for x=t2+1x = t^2 + 1, y=2t3y = 2t^3.

  5. Find the tangent line to x=sec⁡tx = \sec t, y=tan⁡ty = \tan t at t=π4t = \frac{\pi}{4}.

  6. Find dydx\frac{dy}{dx} and d2ydx2\frac{d^2y}{dx^2} for x=ln⁡tx = \ln t, y=t2y = t^2 with t>0t > 0, then confirm by eliminating tt.

Answer to Question 1: dydx=2(t−2)3t2\frac{dy}{dx} = \frac{2(t - 2)}{3t^2}, which equals −23-\frac{2}{3} at t=1t = 1.

Answer to Question 2: dydx=1+t\frac{dy}{dx} = 1 + t and d2ydx2=e−t\frac{d^2y}{dx^2} = e^{-t}.

Answer to Question 3: horizontal tangents at (0,3)(0, 3) and (0,−3)(0, -3); vertical tangents at (4,0)(4, 0) and (−4,0)(-4, 0).

Answer to Question 4: dydx=3t\frac{dy}{dx} = 3t, so d2ydx2=32t\frac{d^2y}{dx^2} = \frac{3}{2t}.

Answer to Question 5: the slope is csc⁡π4=2\csc\frac{\pi}{4} = \sqrt{2} at the point (2,1)(\sqrt{2}, 1), so the tangent is y=2,x−1y = \sqrt{2},x - 1.

Answer to Question 6: dydx=2t2\frac{dy}{dx} = 2t^2 and d2ydx2=4t2\frac{d^2y}{dx^2} = 4t^2; eliminating gives y=e2xy = e^{2x}, with derivatives 2e2x2e^{2x} and 4e2x4e^{2x}, the same values.

Key Takeaways

  • Derivatives of a function in parametric form start from dydx=dy/dtdx/dt\frac{dy}{dx} = \frac{dy/dt}{dx/dt}, valid wherever dxdt≠0\frac{dx}{dt} \neq 0.

  • The formula is the chain rule solved for the slope, and the slope does not change when the curve is traced faster.

  • The second derivative is ddt(dydx)\frac{d}{dt}\left(\frac{dy}{dx}\right) divided by dxdt\frac{dx}{dt}, never the ratio of the two second derivatives.

  • Horizontal tangents come from dydt=0\frac{dy}{dt} = 0, vertical tangents from dxdt=0\frac{dx}{dt} = 0, and points where both vanish need a limit.

  • Circles, ellipses and cycloids are handled without ever writing yy as a function of xx.

Where To Go From Here

Work through Practice Problems 1 to 6; if Problem 2 or 4 trips you up, return to the second derivative section and Mistake 3, and if Problem 3 does, reread the tangent table above Example 1.

Book a free demo class with Bhanzu:

Read More

Was this article helpful?

Your feedback helps us write better content

Frequently Asked Questions

How do you use derivatives of a function in parametric form to find dy/dx for a curve defined by two equations?
Differentiate each equation with respect to the parameter, then divide: dydx=dy/dtdx/dt\frac{dy}{dx} = \frac{dy/dt}{dx/dt}. Simplify the result and keep it in terms of tt.
Find dy/dx for the parametric equations x=2t2−tx = 2t^2 - t, y=t3−2t+1y = t^3 - 2t + 1.
dxdt=4t−1\frac{dx}{dt} = 4t - 1 and dydt=3t2−2\frac{dy}{dt} = 3t^2 - 2, so dydx=3t2−24t−1\frac{dy}{dx} = \frac{3t^2 - 2}{4t - 1}. At t=1t = 1 the slope is 13\frac{1}{3}; at t=14t = \frac{1}{4} the curve has a vertical tangent.
For x=3sin⁡tx = 3\sin t and y=3cos⁡ty = 3\cos t, what is dy/dx in terms of t?
dydx=−3sin⁡t3cos⁡t=−tan⁡t\frac{dy}{dx} = \frac{-3\sin t}{3\cos t} = -\tan t. The curve is a circle of radius 3 traced clockwise, which is why the slope differs in sign from the counterclockwise −cot⁡t-\cot t pattern.
Find dy/dx and d²y/dx² for x=cos⁡(2t)x = \cos(2t), y=cos⁡ty = \cos t, 0<t<π0 < t < \pi.
dydx=−sin⁡t−2sin⁡2t=14cos⁡t\frac{dy}{dx} = \frac{-\sin t}{-2\sin 2t} = \frac{1}{4\cos t}, using sin⁡2t=2sin⁡tcos⁡t\sin 2t = 2\sin t\cos t. Differentiating gives sin⁡t4cos⁡2t\frac{\sin t}{4\cos^2 t}, and dividing by dxdt=−4sin⁡tcos⁡t\frac{dx}{dt} = -4\sin t\cos t gives d2ydx2=−116cos⁡3t\frac{d^2y}{dx^2} = -\frac{1}{16\cos^3 t}, valid for t≠π2t \neq \frac{\pi}{2}.
A curve has x=t2x = t^2 and y=sin⁡(t2)y = \sin(t^2). What is dy/dx?
dydx=2tcos⁡(t2)2t=cos⁡(t2)\frac{dy}{dx} = \frac{2t\cos(t^2)}{2t} = \cos(t^2) for t≠0t \neq 0. That matches eliminating the parameter: the curve is y=sin⁡xy = \sin x for x≥0x \geq 0, and cos⁡x=cos⁡(t2)\cos x = \cos(t^2).
How do you find higher-order derivatives of a function in parametric form?
Repeat the second-derivative move. Each new derivative is ddt\frac{d}{dt} of the previous one, divided by dxdt\frac{dx}{dt}. For x=etx = e^t, y=e3ty = e^{3t} this gives 3e2t3e^{2t}, then 6et6e^{t}, then 66, matching y=x3y = x^3.
Is the parametric second derivative ever equal to the ratio of the second derivatives?
Yes, but only in special cases, such as when d2xdt2=0\frac{d^2x}{dt^2} = 0 and dxdt=1\frac{dx}{dt} = 1 (for example x=tx = t). In general the ratio is wrong, as Example 3 shows on the circle.
✍️ Written By
BT
Bhanzu Team
Content Creator and Editor
Bhanzu’s editorial team, known as Team Bhanzu, is made up of experienced educators, curriculum experts, content strategists, and fact-checkers dedicated to making math simple and engaging for learners worldwide. Every article and resource is carefully researched, thoughtfully structured, and rigorously reviewed to ensure accuracy, clarity, and real-world relevance. We understand that building strong math foundations can raise questions for students and parents alike. That’s why Team Bhanzu focuses on delivering practical insights, concept-driven explanations, and trustworthy guidance-empowering learners to develop confidence, speed, and a lifelong love for mathematics.
Related Articles
Book a FREE Demo ClassBook Now →