Mothers and fathers prefer different games when playing with their children, nonetheless, they share common latent play characteristics. This study aimed to evaluate an observational coding system comprising three rating scales for adjustment, calibration, and familiarity of parent-child play situations that, together, indicate latent parent-child play-quality. 335 toddlers and preschoolers and their parents (332 fathers and 158 mothers) were observed during both physically (father-preferred) and cognitively stimulating (mother-preferred) parent-child games. Measurement, structural, and functional invariance were assessed concerning the type of game and parent gender. No differences were found in any respect. Results showed that the coding system is reliable and unbiased for both parents and game types, indicated equal validity in regard to the external criterion attachment security, and revealed no differences in latent parent-child play-quality across parent gender. © 2018 The Author(s). Published by Informa UK Limited, trading as Taylor & Francis Group