Exam FAM Sample Questions, Free and Interactive
FAM is two exams that share a sitting. The short-term half (loss models, credibility, frequency/severity) and the long-term half (survival models, life insurance pricing, multi-state models) test fundamentally different mental models. Candidates who treat FAM as one exam and try to alternate concepts day-by-day routinely underperform. Candidates who treat it as two parallel study tracks tend to pass.
The SOA's 262 published FAM sample questions are split roughly evenly across the two halves. FreeFellow hosts all 262 inside the practice surface with worked solutions alongside, performance tracking by topic, and an adaptive engine that re-surfaces missed items. Free, no signup to browse, no credit card.
Start practicing Exam FAM samples
What the 262 Sample Items Cover
The split between halves and the topic emphasis within each:
| Half | Topic area | Approx. count |
|---|---|---|
| Short-term (FAM-S) | Frequency models (Poisson, negative binomial, binomial) | ~25 |
| Short-term | Severity models (lognormal, Pareto, Weibull, exponential, gamma) | ~30 |
| Short-term | Coverage modifications (deductible, limit, coinsurance, inflation adjustments) | ~30 |
| Short-term | Aggregate loss models (compound Poisson, Panjer recursion, stop-loss) | ~30 |
| Short-term | Credibility (limited fluctuation, Bühlmann, Bühlmann-Straub) | ~25 |
| Long-term (FAM-L) | Survival models (force of mortality, life table mechanics) | ~35 |
| Long-term | Life insurance pricing (whole life, term, endowment) | ~25 |
| Long-term | Annuities (life-contingent, joint-life, status-based) | ~20 |
| Long-term | Reserves (net premium, full preliminary term) | ~15 |
| Long-term | Multi-state Markov models | ~27 |
The recognizable patterns:
Aggregate loss problems almost always involve a compound Poisson with a specific severity distribution; the candidate is asked for the variance of S, the stop-loss premium at a fixed retention, or the probability of S exceeding a threshold. Bühlmann credibility problems load heavily on extracting n from a data structure that obscures it (a credibility-weighted estimator combined with a hypothetical mean). Multi-state model problems favor three-state setups (alive, disabled, dead) where the candidate computes a transition probability or a discounted expected value.
Why FAM Samples Earn Their Reputation
FAM is the bridge exam. P and FM test math and finance; the FSA-track exams test deep actuarial work; FAM is the first exam where you have to think like an actuary, not a mathematician. The samples train exactly this shift.
The SOA samples force you to confront the recurring traps unique to FAM: a Poisson rate that varies by class with unequal class sizes (Bühlmann-Straub setup), a survival function with a piecewise hazard that flips at age 80, a multi-state model where the candidate has to recognize that transitions are time-homogeneous before applying the constant-force shortcut.
FAM has more "did you read the stem carefully" questions than any other preliminary exam. Inflation adjustments to deductibles, coverage modifications applied per-claim vs. per-policy-year, mortality assumptions that change at a specific age. The SOA writes these to catch candidates who pattern-match instead of reading. Sample-question practice trains the careful-reading reflex.
Three Tactics for the 262 Items
Study the two halves in parallel, not in series. Alternate study weeks between FAM-S and FAM-L rather than trying to learn one half completely before starting the other. Both halves take time to "click" and the click happens during sleep consolidation. Parallel tracks make better use of the consolidation calendar.
Drill Bühlmann to reflex. The Bühlmann credibility weight Z = n / (n + k), the Bühlmann-Straub generalization with class-size weighting, and the partial-credibility predictor structure all combine into a single recognizable problem shape that the SOA returns to repeatedly. Drill the formula, then drill the data-structure recognition (extracting n from a "we observed average claim severity of X over Y class-years" stem).
Derive every life insurance formula from first principles once. Then memorize. The shortcut formulas for whole life, term insurance, endowment insurance, and the various status notations all derive from the same expected-value-of-discounted-payoff integral. Doing the derivation once builds the intuition; memorization gives you the speed.
What the Samples Don't Cover Well
Honest gaps for FAM:
- Multi-state models with four or more states. The sample set leans on three-state setups. Live exam can deploy four-state models (e.g., healthy, mild disability, severe disability, dead).
- Frequency-severity dependent models. When frequency and severity aren't independent (a real-world commonplace), the SOA samples are light. Supplement with textbook exercises (Klugman/Panjer/Willmot).
- Pension-style joint-life with reversions. Light coverage; supplement from the FreeFellow original FAM bank.
How FAM Practice Compares Across Free and Paid
| Source | Exam FAM Sample Questions | Format | Cost |
|---|---|---|---|
| FreeFellow | All 262, interactive | Same surface as topic practice | $0 |
| ASM FAM Manual | All 262 in study manual | $130 to $200 | |
| ACTEX FAM Manual | Selected items + commentary | $200 to $300 | |
| Coaching Actuaries Learn + Adapt FAM | All 262 + originals + EL scoring + video lessons | Interactive within Adapt | $300 to $400 per exam |
| SOA only | All 262 | Static PDF on soa.org | $0 |
FAM is the most expensive preliminary exam to prep through commercial providers because the conceptual depth (and the two-halves split) traditionally justifies bigger paid bundles.
Start Practicing
Practice all 262 Exam FAM samples. For the broader FAM study plan with weekly milestones: How to Pass Exam FAM in 4 Months. For ASTAM (the constructed-response successor that builds on FAM): Free Exam ASTAM Sample Questions.