The "conditions" are circuit dependent. Sometimes there may not even be a maximum (such as a single capacitor, whose impedance goes to infinity as frequency goes to zero). The most common way to analyze this is to treat the circuit as a 2 port model, with an input voltage between two wires and an output voltage between two other wires. Once we think about the circuit this way, the maximum impedance is the frequency where the gain of this circuit is at a minimum.
You can do this by hand in the time domain, but usually we take advantage of the fact that the LCR circuit is linear and use the Laplace transform which converts the time domain description of this circuit (with all its derivatives and integrals) into an algebraic formula in terms of a "complex frequency" (usually denoted with $s$). This is very convenient because finding the minimum of an algebraic equation is much simpler than finding the minimum of a complicated mix of derivatives. One finds the $omega$ such that when one substitutes $[s:= j\omega]$ into that algebraic expression, it's a minimum. That frequency is the frequency where the minimum occurs, and the resulting gain can be used to compute the impedance at that point.
Alternatively, if you haven't learned Laplace transforms yet, it may be easiest to put it into a circuit analyzer and just have it plot the response at different frequencies. But if you want to do it by hand, Laplace transforms are the tool of choice to do it as easily as possible.