**Background:** The period from pregnancy through 2 years of life is a critical window for preventing childhood malnutrition, with complementary feeding covering the widest part of this window. Despite Ethiopia's infant and young child feeding (IYCF) guidelines, sub-optimal complementary feeding remains a major contributor to inadequate nutrient intake among children aged 6–23 months. Previous studies in Ethiopia have used one or two indicators inconsistently, and no evidence existed on optimal complementary feeding practices across different agro-ecological zones in the Jimma zone or Oromia region. This study aimed to determine optimal complementary feeding practices and associated factors among children aged 6–23 months in three agro-ecological rural districts (high, mid, and lowland) of Jimma zone, Ethiopia.
**Methods:** A community-based cross-sectional study was conducted from February to March 2020 in three purposively selected districts: Dedo (highland), Mana (lowland), and Kersa (midland). The sample size of 845 mother-child pairs was calculated using a single population proportion formula (95% confidence level, 5% precision, 50.9% proportion from a previous study, design effect of 2, and 10% non-response rate). Multistage sampling was employed: three districts were selected purposively, then kebeles were selected by simple random sampling (7 from Dedo, 5 from Mana, 5 from Kersa), and households were selected via systematic sampling. Data were collected using structured, pretested questionnaires translated into Afan Oromo. Dietary data were collected using a 24-hour recall method. According to WHO 2021 guidelines, MDD was computed from eight food groups, MMF was based on feeding frequency per 24 hours, and MAD was a composite indicator of both MDD and MMF. Timely introduction was defined as receiving solid, semi-solid, or soft foods at 6 months. OCFP required meeting all four indicators. Maternal knowledge was assessed from nine questions, with a score of ≥4 (mean 3.7±1.56) classified as good knowledge. Data were entered into Epi Data V.1.4.4.0 and analyzed using SPSS version 20. Binary and multivariable logistic regressions identified factors associated with OCFP, with significance at p<0.05. Multicollinearity was checked (standard error <2.0) and model fitness was assessed by the Hosmer-Lemeshow test (p=0.781).
**Key Results:** Of 845 mother-child pairs, 843 (99.3%) provided complete responses. The mean child age was 13±4.7 months, 44.5% were aged 6–11 months, and 52% were male. The mean maternal age was 26.8±5 years, 54.2% were aged 25–30 years, 92.5% were married, 95.8% were Oromo, and 96.6% were Muslim. Over half of mothers (63%) and fathers (48.6%) were unable to read or write, and 91.9% of mothers were housewives. The average family size was six. The overall proportion of OCFP was 9.4% (95% CI: 7.19, 11.08). Timely initiation of complementary feeding was 52.2%, MMF was 64.1%, MDD was 17.2%, and MAD was 12.2%. Across agro-ecological zones, OCFP was 16.0% in highland, 12.6% in lowland, and 4.4% in midland (p<0.001). MDD was highest in highland (23.5%) and lowest in midland (8.4%) (p<0.001). MMF was highest in lowland (73.6%) (p<0.05). MAD was 17.2% in highland, 16.1% in lowland, and 6.3% in midland (p<0.001). The most commonly consumed food group was grains, roots, and tubers (92.5%), while flesh foods were least consumed (14%). Vitamin A-rich fruits and vegetables were consumed by 27.5% of children, and other fruits and vegetables by 20.8%. In highland, 39.9% consumed vitamin A-rich fruits and vegetables (p<0.001) and 28.2% consumed other fruits and vegetables (p<0.05). Flesh food consumption was highest in lowland (35.1%) (p<0.001). Overall, 42.7% of mothers had good knowledge about complementary feeding, with the highest proportion in lowland (72.4%) and lowest in midland (33.6%) (p<0.001). Multivariable logistic regression showed that living in highland (AOR=4.215, 95% CI: 2.204, 8.062), good maternal knowledge (AOR=3.952, 95% CI: 2.121, 7.366), maternal primary education (AOR=2.817, 95% CI: 1.516, 5.235), and family size of 1–3 persons (AOR=3.649, 95% CI: 2.116, 6.292) were positively associated with OCFP.
**Clinical Implications:** The extremely low prevalence of optimal complementary feeding (9.4%) indicates that the vast majority of children aged 6–23 months in rural southwest Ethiopia are at high risk of undernutrition during the critical window of development. The marked variation across agro-ecological zones—with midland areas performing worst—suggests that agricultural production diversity directly influences dietary diversity and child feeding quality. The strong associations with maternal education and knowledge highlight the need for targeted nutrition education interventions, particularly for mothers with no formal schooling. The inverse relationship with family size suggests that family planning services could indirectly improve child feeding outcomes. Multi-sectoral collaboration involving health, agriculture, and education sectors is needed to strengthen maternal education, promote diversified agricultural production (especially in cereal-dependent midland areas), and improve access to nutrition information. The low consumption of animal-source foods (14%) and vitamin A-rich fruits and vegetables (27.5%) across all zones indicates a need for interventions promoting dietary diversity beyond staple grains.