The sudden switch to online learning brought on by the COVID-19 pandemic has completely transformed the way we teach and evaluate English language proficiency. It has put traditional assessment methods, which were designed for controlled, in-person settings, to the test in ways they were never meant to handle. This study dives into the effectiveness, challenges, and equity issues surrounding online English language assessment, using a convergent mixed-methods approach. We combined a quantitative survey involving 250 educators and students with semi-structured qualitative interviews from 15 educators and assessment experts. Surprisingly, participants rated online assessments as somewhat more effective than a neutral baseline (M = 3.42, SD = 1.05, t = 6.32, p<.001), although this positive feedback wasn’t evenly spread out. Access to technology turned out to be the strongest and most consistent factor influencing both perceived effectiveness and equity concerns (r = .52 and r = -.61 respectively, both p<.001), with significant regional differences showing that North America and Europe had the upper hand compared to South Asia and Sub-Saharan Africa. Opinions on AI-driven grading tools were mixed, landing close to neutral, and perceptions of their accuracy were inversely related to concerns about bias (r = -.41, p<.001). On the other hand, formative assessment practices were the most positively received innovation, with 75.6% of respondents finding them helpful or very helpful. These results suggest that the future of online English language assessment hinges less on the complexity of any one technology and more on bridging the technological access gap that dictates who gets to benefit from it. The paper wraps up with practical, evidence-based recommendations for educators, institutions, and policymakers.