Research needs to be peer reviewed transparently so readers may have more confidence on objective and unbiased peer review, and consequently more trust in the accuracy of the published research studies. In spite of the numerous critics peer review still serves as the primary quality assurance system in scholarly journals. The objective of this study was to identify whether there is a need to develop instructions for peer reviewers which will make peer review process more transparent and instructions more helpful to reviewers. We conducted a content analysis of 84 available instructions for peer reviewers from the HRČAK portal for Croatian Open Access (OA) journals with more than 350 OA journals at the present. A non-validated categorization scheme for coding text was developed containing 3 main categories, information about reviewer, peer review, and submitted manuscript. In order to cover each category well 11 subcategories were added at first, 18 at second, and 2 at third hierarchical level. Subcategories included in total 269 words, phrases and rules. Coding unit (case) was a document in PDF or DOC format, containing instructions for reviewers for the specific journal in English or Croatian language. Results were expressed as frequencies, case occurrences, and percentages There is limited guidance and no consensus regarding the optimal content and structure of the instructions for peer reviewers. Among 84 instructions of Croatian OA journals, 64 are just reviewer forms without explanations and meaning of stated questions, or instructions in which way should reviewers provide answers to them. The most present category was information about manuscript (83/84), with manuscript elements (title, abstract, introduction, methods, discussion, conclusion, and acknowledgement) as most frequent subcategory (82/84). Research data (raw data, underlying data) were not mentioned in a single instruction. Information about reviewer was present in the majority of instructions (79/84), with reviewers#39; comments and suggestions as most frequent terms (50/84 and 45/84 accordingly). It was surprise that peer review was the least represented category in the instructions for peer reviewers (71/84). Among peer review subcategories the most present was about revision results (accepted, rejected...)(66/84), while subcategories peer review types (blind, anonymous, open...), peer review process (confidentiality, fairness, unbiasedness...) and ethical issues (authorship, misconduct, redundancy, plagiarism...) were represented poorly (22-28/84). Croatian OA journals are not recognizing the importance of peer review and the transparency of the whole process. Among approximately 350 Croatian OA journals we found only 84 instructions for reviewers in English or Croatian language. The majority of key peer review issues are poorly reported in the survey of instructions for peer reviewers. Peer review process in the majority of the Croatian OA journals is not transparent. Also open peer review is not accepted by Croatian journals. Our findings highlight the need for raising awareness about the importance of transparency of peer review, and clear and consistent peer review guidelines.