This study examined the relationships between the perceived phonological salience of Nigerian English-language fiction, reading performance and academic self-efficacy among senior secondary school students in Owerri, Nigeria. A correlational design was adopted; 120 senior secondary two (SS2) students were drawn from four public co-educational senior secondary schools through multistage sampling. Eight passages from four Nigerian novels were scored for alliteration, rhyme, rhythm and sound repetition by three raters (Fleiss' κ = .87). All students read the same four passages, selected to span the density range, and rated the salience of their sound patterning on a 16-item scale yielding four subscales and a composite. Reading performance was assessed through oral reading rate (words read correctly per minute) and oral reading accuracy on an unpracticed passage, and through a 20-item passage-based comprehension test; academic self-efficacy was measured with a 20-item questionnaire. Data were analyzed using descriptive statistics, Pearson's product-moment correlation and Fisher-z confidence intervals at the .05 level. Composite salience scores correlated positively with oral reading rate (r = .31, p = .001, 95% CI [.14, .46]), text comprehension (r = .27, p = .003, [.10, .43]) and academic self-efficacy (r = .22, p = .016, [.04, .38]), and oral reading rate correlated strongly with comprehension (r = .58, p < .001, [.45, .69]). All four null hypotheses were rejected. Rhythm was the numerically strongest correlate of every outcome, but Steiger's tests showed that no pairwise difference between features reached significance, and power for that question was approximately 27%. Because the design was correlational, the findings indicate association rather than causation; the single-zone sample, the use of one oral reading passage, the self-report predictor and unmodelled school-level clustering limit inference. The study recommends that English language teachers incorporate