Private general undergraduate institutions (PGUIs) account for 31.5% of undergraduate institutions in China, yet the vocabulary learning needs of their English majors remain underexplored, highlighting a critical research gap. This study adopts Hutchinson and Waters’ (1987) needs analysis framework to examine PGUI students’ Test for English Majors Band-4 (TEM-4) vocabulary learning needs. Employing a convergent parallel mixed-methods design, this study integrated questionnaire data from 240 students with semi-structured interviews from seven experienced teachers. The findings indicated that 77.9% of students recognized vocabulary as essential for TEM-4 performance, reflected in a high mean score for perceived necessity (M = 3.98, SD = 1.05). However, they reported moderate perceived lacks (M = 3.62) in current learning approaches. Students showed strong acceptance of mobile-assisted learning (M = 3.80) and high expectations for deep learning functions (M = 3.71). Qualitative analysis of 475 codes across 26 sub-themes revealed that teachers identified a critical 3:1 receptive-productive gap, with students recognizing approximately 6,000 words but actively using fewer than 2,000. All seven teachers confirmed that existing mobile applications, despite widespread use, fail to support deep vocabulary learning outcomes. The integration of quantitative and qualitative findings provided clear guidelines for developing TEM-4-specific mobile applications, highlighting systematic testing with immediate feedback, authentic TEM-4 examples, vocabulary discrimination exercises, and multimodal learning resources. The study provides empirical evidence for designing mobile applications that address the multidimensional vocabulary learning needs of PGUI students, a distinct and under-researched learner population.