Purpose: Phonological awareness (PA), the awareness of sounds in spoken words, is strongly linked to reading outcomes. However, there is an ongoing debate regarding the effectiveness of PA instruction without including print (i.e. PA without exposure to words or letters). Specifically, is PA-only instruction just as effective in improving reading outcomes when compared to print-based instruction (e.g. phonics)? Method: Robust variance estimation was used to examine the effectiveness of PA-only instruction across 38 independent studies and 293 effect sizes. The sample consisted of 3,880 participants in preschool through first grade. The average age of participants was 5 years 7 months. Moderator analysis were conducted to examine differences in effect across outcomes, comparison conditions and participant characteristics. Results: PA-only instruction had a moderate overall effect (g = 0.57, 95% CI [0.40, 0.74], p < 0.001) on reading-related outcomes. PA-only instruction was less effective when compared to print-based instruction (g = -0.54, 95% CI [-0.80, -0.27], p < 0.001) and less effective for word reading outcomes (g = -0.33, 95% CI [-0.66, -0.01], p = 0.05) than for PA outcomes (g = 0.64, 95% CI [0.44, 0.84], p < 0.001). Conclusion: PA-only instruction can be beneficial for reading-related outcomes, but the effect varies depending on two important contextual factors: 1) the reading outcome type and 2) the comparison condition. Specifically, PA-only instruction might be less effective in improving word reading outcomes than PA outcomes. In addition, it has less of an effect on reading-related outcomes when compared to print-based instruction like phonics.