Rubisco is the enzyme responsible for the net photosynthetic carbon fixation in plants. The enzyme consists of chloroplast-encoded large subunits and a family of nuclear-encoded small subunits (SSUs). SSUs are essential for the assembly of the Rubisco enzyme in higher plants and have been shown to influence catalytic activities of Rubisco (Atkinson et al., 2017; Laterre et al., 2017). This study aims to characterise the roles of SSUs in the model plant Arabidopsis thaliana (Arabidopsis). Arabidopsis contains four SSU (RbcS) genes, RbcS1A (1A hereafter) on chromosome 1, and RbcS1B, RbcS2B and RbcS3B (1B, 2B, 3B, respectively hereafter) located in tandem on chromosome 5. To characterise the roles of SSU, I have i) generated SSU knock-out mutants using CRISPR/Cas9 editing and analyse the impact of single and multiple SSU knock-outs on plant growth and fitness; ii) measured RbcS expression under different light qualities, quantities and temperatures; iii) performed growth analyses under different environmental conditions using CRISPR/Cas9 rbcs mutants; and iv) complemented the triple mutant 1a2b3b (BigBoi) with SSU from Chlamydomonas reinhardtii (Chlamydomonas) and performed growth analyses on the complemented lines. To generate SSU knock-out mutants via CRISPR/Cas9, two pairs of gRNAs were designed to target each RbcS gene. The efficiency of the gRNAs pairs was evaluated in mesophyll protoplasts and it was found that at least one pair for each gene was able to induce large deletions of 96-180 bp in RbcS. Constructs containing Cas9 nuclease and gRNAs were stably transformed in Arabidopsis by floral dipping. The T1 progeny containing the transgene were screened for large deletions by PCR and small insertions/deletions (indels) by Sanger sequencing. PCR screening showed that large deletions were induced in planta, but they were chimeric deletions and not transmittable to the T2 progeny. On the other hand, indels of 1-7 bp occurred at a higher rate (12% and 7%, respectively) and were heritable. Analyses of the T2 progeny revealed that heterozygous mutation was the most common type of mutation in 1B and 2B but chimeric mutation was the most common in 1A and 3B. The heritability rates for 1a, 1b, 2b and 3b were 4%, 20%, 8% and 6%, respectively. To measure the RbcS expression under different environmental conditions, the arrhythmic clock mutant prr5/7/9 mutant was grown under constant light for 14 d. The mutant was kept in darkness for 24 h before exposing to white light for 12 h. Transcript analysis was performed and the result showed that i) all RbcS genes were induced by light and the total transcript abundance increased when light was turned on and decreased after light was turned off; and ii) each RbcS had different induction and degradation rates. 1A was induced the most quickly and degraded most rapidly while 1B was induced the most slowly and 2B was the most stable transcript after light was turned off. The experiment was repeated but with different light qualities (blue, red, and far-red), light quantities (high (1000 μmol photon m-2 s-1), medium (200 μmol photon m-2 s-1) and low (50 μmol m-2 s- 1) light), and temperatures (high (30oC) and low (10oC), white light at 200 μmol m-2 s-1). Transcript analyses showed that blue light induced the highest level of increase among light qualities followed by red light and far-red light. High light induced the highest level of transcript abundance followed by medium light and low light. Under high temperature, the expression of 2B and 3B increased significantly and 3B was the major isoform. On the other hand, the expression of 2B and 3B were suppressed under low temperature and 1A was the major isoform. This suggested that 2B and 3B were the most sensitive temperature mediators of the RbcS gene family. Based on the transcript analyses, the high light, high temperature and low temperature conditions were chosen for growth experiments. WT, T-DNA and CRISPR/Cas9 mutants were used to test the following hypotheses: i) under the light saturating condition (high light), Rubisco becomes limiting and plants with reduced Rubisco content (1a, 3b, 2b3b and 1a2b) would grow more slowly than WT; ii) under high temperature where 3B is the major isoform, plants lacking 3B (3b, 2b3b) would suffer a reduced growth rate relative to WT; iii) under low temperature where 1A is the major isoform, plants lacking 1A (1a, 1a2b) would suffer a reduced growth rate relative to WT. Growth under high light was able to differentiate the areas of 1a, 2b3b and 1a2b mutant in comparison to WT, but not 3b mutant. However, the weight of 3b mutant was significantly lower than that of WT, suggesting the leaves of 3b were thinner. Under high temperature, 3b and 2b3b mutants were not significantly different from WT. This was due to 3B accounting for ca 50% of the transcript abundance under high temperature and growth was found to be unaffected at this level of RbcS decrease. Under low temperature, the areas of 1a and 1a2b were not significantly different from that of WT, but weights were significantly lower. This was similar to 3b under high light and suggested that leaves of plants with significant Rubisco reduction become thinner first and further decrease in Rubisco resulted in the loss of leaf area. Altogether, this study showed that RbcS genes collectively contribute to the overall transcript abundance and 2B and 3B genes are most susceptible to the changes in temperature. The triple mutant 1a2b3b generated in this study was used as a model to study the effects of heterologous SSUs to growth. After complementation with Chlamydomonas SSU, seven independent complemented lines were identified and the slow-growing phenotype was rescued in all lines. The area of complemented plants ranged from 8-34% of WT compared to 1% of the triple mutant on day 28. This study showed that the triple mutant could be used as an Arabidopsis platform to study the effects of heterologous SSUs to Rubisco catalytic activities.