Conflicting evidence exists on how elk populations grow. Two models, Ricker and θ‐logistic, are used to study this. A simulation showed that when vital rates vary a lot, both models fit similarly. The θ‐logistic model is more realistic when vital rates vary little and growth is nonlinear. But it struggles with high variation and vague priors. For elk populations with slow life histories, the Ricker model is usually better, especially with high environmental variation or vague priors.