We examined whether mothers' use of temporal synchrony between spoken words and moving objects, and infants' attention to object naming, predict infants' learning of word–object relations. Following 5 min of free play, 24 mothers taught their 6‐ to 8‐month‐olds the names of 2 toy objects, Gow and Chi, during a 3‐min play episode. Infants were then tested for their word mapping. The videotaped episodes were coded for mothers' object naming and infants' attention to different naming types. Results indicated that mothers' use of temporal synchrony and infants' attention during play covaried with infants' word‐mapping ability. Specifically, infants who switched eye gaze from mother to object most frequently during naming learned the word–object relations. The findings suggest that maternal naming and infants' word‐mapping abilities are bidirectionally related. Variability in infants' attention to maternal multimodal naming explains the variability in early lexical‐mapping development.