Purpose and Rationale:Numerous students experience difficulties in recalling glyphs (e.g., strokes, radicals, and spatial composition forms) when they are learning to write Chinese characters, and traditional copying (TC) exercises do not seem to be of much help to these children. Studies have indicated that a main cause of Chinese handwriting difficulties is an inability to recall Chinese character glyphs. Accordingly, an intervention focused on recall of Chinese character glyphs may be an effective means of enhancing the learning of Chinese character writing. Furthermore, experimental studies have demonstrated that long-term memory is improved when part of the learning process is focused on active information retrieval. Scholars have indicated that during the learning process, repeated retrieval practice has a greater effect on learning outcomes than repeated studying does. The retrieval effect, also referred to as the testing effect or active recall, is the focus of the present study, which investigated whether forced-recall handwriting instruction (FR method), a teaching method that emphasizes retrieving Chinese character glyphs from memory, leads to more favorable learning outcomes relative to the TC method. The FR method requires students to read out the component radicals of target Chinese characters and to perform the “write a character, then cover the character” procedure, in which students have to cover the just-learned target Chinese character with hand and recall the glyphs of the character, then reproduce it on the workbook without a printed target character to refer to. By contrast, the TC method requires students to copy a printed target character and perform monotonous and repetitive copying exercises without additional deliberate effort. Methods: Two third-grade male students who were from middle-class families, exhibited normal intelligence, had handwriting difficulties, and were from the same class of a school participated in the present study. Notably, their final scores for their Mandarin course consistently placed them in the bottom 20% of their class. In the Comprehensive Test of Basic Chinese Character Reading and Writing (Hung et al., 2003), they scored in the 10th and 6th percentiles on the “character writing based on phonetic-symbol-spelled words” subtest and in the 6th and 1st percentiles on the “dictation” subtest. However, their percentile rankings in the Chinese Character Recognition Scale for Elementary Graders were 37th and 31st, within the normal range. The two participants had average pen-holding abilities, copying speeds, and literacy levels, and they did not exhibit any physical or mental disability or experience any cognitive-motor coordination difficulties that could explain their Chinese handwriting difficulties. Although the two participants exhibited average character-sizing and reading abilities during Mandarin classes, they often encountered difficulties in recalling Chinese character glyphs and using phonetic symbols to replace target Chinese characters. Because they frequently committed writing errors, such as confusing Chinese characters with similar pronunciations or appearances, misplacing radicals, adding extraneous strokes, and omitting required strokes, their regular Mandarin assessments consistently placed them in the bottom 20% of a class of 25 students. In the present study, a single-subject alternating intervention design was adopted. The implemented intervention comprised 12 40-min sessions, with 2 sessions being completed per week. These sessions were conducted by a single teacher, who taught both students simultaneously during each session. For the alternating treatment phase, eight sessions were completed, with each session incorporating both FR and TC trials. To minimize potential biases due to the order of the trials, we conducted the TC trial before the FR trial in odd-numbered sessions and the FR trial before the TC trial in even-numbered sessions. All target Chinese characters selected for the study were characters that the participants had yet to learn. The researchers of the present study controlled for the number of strokes of the target characters learned under both teaching conditions. In the final phase, comprised of 4 25-min sessions, only the FR method was implemented. To evaluate learning outcomes, this study collected data on immediate, 1-week, and 2-week retention by conducting a curriculum-based assessment of the participants’ Chinese character writing skills in which they were required to reproduce specific target characters from memory without referring to printed target characters. Results/Findings: The results of the present study indicate that the FR method led to higher retention rates relative to those achieved using the TC method. For immediate, 1-week, and 2-week retention, the two participants achieved consistently higher average character retention rates through the FR method than through the TC method. Wilcoxon signed-rank tests were conducted to compare the retention outcomes of the two methods, and four out of five comparisons revealed significant differences in the retention rates between the two methods at a significance level of .05. Our qualitative results also indicate that the FR method was simple and effective and that it was well-received by the participating teacher and students. Conclusions/Implications: First, our findings support that the FR method can be implemented in place of the TC method for teaching Chinese character handwriting because of its more favorable learning outcomes. When Chinese character glyphs are complex, lower-grade elementary students are most frequently assigned character copying exercises as homework. Copying new characters is a time-consuming daily activity that every student learning Chinese experiences during the first few years of their schooling life. Our results indicate that the FR method leads to more favorable learning outcomes relative to the TC method. Moreover, the FR method does not add to the workload of students. Our findings can serve as a useful reference for teaching new Chinese characters to lower-grade elementary school students. Second, we contribute novel insights to the literature on information retrieval. Most related studies have used spoken- or written-language stimuli as learning materials in their FR trials. For example, a list of words in Swahili was used in a retrieval experiment conducted by Karpicke and Roediger (2008). In that study, participants were asked to provide oral responses (e.g., speaking the English translation of a target Swahili word). By contrast, the present study did not focus on spoken output; rather, it focused on mental representations of outcome measures related to Chinese character writing, which are different from those associated with spoken language. Furthermore, in the context of linguistic and phonological information processing, writing a Chinese character requires visual-spatial-motor coordination. Our findings indicate that the FR method can be applied to materials other than those focused on spoken or written language. Further studies on the effects of retrieval practice should explore multiple learning modes to expand on our findings.