Reading comprehension is contingent on both oral language comprehension and word-level reading ability, skills that are thought to be intrinsically related in the early school years. However, while previous studies examining bidirectional relationships among oral vocabulary and reading development have generally found an association between word recognition and subsequent vocabulary skills, they have yielded inconsistent results regarding whether vocabulary is linked to later word recognition. In this longitudinal study (n = 176), we assessed reciprocal relationships between expressive and receptive vocabulary and word recognition measured yearly from Grade 1 to 5. We compared results from a cross-lagged panel model (CLPM) to a random-intercept cross-lagged panel model (RI-CLPM), an analytic approach which permits disaggregation of variance between children in the constructs of interest from year-to-year fluctuations within children. Expressive and receptive vocabulary were examined in separate models. Results showed that the RI-CLPMs provided a better fit to the data in contrast to the CLPMs, and provided evidence of reciprocal relationships between word recognition and both receptive and expressive vocabulary. Significant between-child variance was shown for vocabulary but not word recognition. These findings confirm that reading and vocabulary skills build upon one another across the primary years, and each is an important area of classroom instruction focus. Early and regular screening measures may serve to inform the development of timely and appropriate support and intervention.