<p>Acoustic spatial capture-recapture (ASCR) surveys with an array of synchronized acoustic detectors can be an effective way of estimating animal density or call density. However, constructing the capture histories required for ASCR analysis is challenging, as identifying which detections at different detectors are of which calls is not a trivial task. The time delays caused by varying distances between call source locations and detectors mean that the order of detection may not reflect the order of emission, and without resolving call identities, the number of distinct detected calls remains unknown. We propose a Monte Carlo Expectation Maximization (MCEM) method to address this unknown call identity problem. In the expectation step, we sample latent variables from a complete-data likelihood model, and in the maximization step, we use either a semi-complete-data likelihood or a conditional likelihood to estimate model parameters. Confidence intervals are obtained via parametric bootstrap. Applied to a survey of moss frogs, our method produces a call density estimate within 15% of that obtained using manually constructed capture histories. Unlike the manual approach, however, our confidence interval incorporates the uncertainty in detection matching. Simulations further demonstrate a low bias (around 6%) and confidence interval coverages close to the nominal 95% level.</p>